Download Is the composition of public expenditures converging in EMU countries

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Economic growth wikipedia , lookup

Business cycle wikipedia , lookup

Fiscal multiplier wikipedia , lookup

American School (economics) wikipedia , lookup

Economics of fascism wikipedia , lookup

Stability and Growth Pact wikipedia , lookup

Transcript
An analysis of the convergence of the composition of public
expenditures in EU countries
Jesus Ferreiro
M. Teresa Garcia-del-Valle
Carmen Gomez
Abstract:
The economic literature on fiscal policies is paying increasing attention to the impact of
the composition of public expenditures on long-term economic growth. Public policy
endogenous growth models recommend to change the composition of public
expenditures to items considered as productive expenditures. Based on these models,
European institutions are encouraging the rise in the share of some of these outlays, like
public investments, R&D, active labour market policies, etc. The paper analyses
whether these recommendations are followed by EU countries and whether a
convergence to a new pattern of public finances with a higher share of productive
expenditures in these countries is arising in the European Union.
Key words: Public expenditures, European Union, convergence, quality of public
finances
1
An analysis of the convergence of the composition of public
expenditures in EU countries
1. Introduction
The European Monetary Union (EMU) approach to economic policy requires its
Member states to implement an orthodox strategy of macroeconomic policy that gives a
special role to monetary policy, while at the same time downgrades fiscal policy. Fiscal
policy in EMU is determined by the Maastricht Treaty and the Stability and Growth
Pact (SGP); these require national fiscal policies in the eurozone and in candidate
countries to avoid excessive fiscal deficits (fiscal deficits below 3% GDP and stocks of
public debt below 60% GDP) and, simultaneously, to reduce the size of public sectors.
These principles are based on a theoretical background according to which fiscal policy
cannot affect the economic activity in the long-run (as measured by potential output).
An active fiscal policy can only be implemented on a short-term basis, correcting
cyclical disequilibria through the working of built-in stabilizers. The long-term effects
would arise from the so-called non-Keynesian effects of fiscal policy. The literature
about the non-Keynesian effects of fiscal policy has given rise to the development of a
number of studies focusing on the expansionary impact of fiscal consolidation.
For the orthodox view, demand-side policies do not affect economic activity in the long
run, that is, fiscal policy does not influence the path of potential output. However,
recently fiscal policy is gaining relevance. The Lisbon Strategy, the current Broad
Economic Policy Guidelines (BEPG) and the reformed SGP accept the potential
positive impact that fiscal policy can have on economic activity in the long-run, both in
terms of the level and the long-term rate of growth of the potential output. This impact
would come from the composition of public expenditures. Thus, the share of
‘productive’ expenditures should be increased.
This strategy should have generated a convergence in national fiscal policies, both in the
sign (and size) of fiscal imbalances, in the size of public revenues and spending and in
the composition of these items. Fiscal rules arising from the Maastricht Treaty and the
2
Stability and Growth Pact seem to have led to a convergence of public imbalances, the
size of public deficits and the stock of public debt, but the outcome is not so evident in
the case of the size of public sectors and in the composition of public expenditures. In
the latter case, although some studies support the hypothesis of convergence in the
composition of public expenditures (European Commission, Directorate-General for
Economic and Financial Affairs, 2002; Sanz and Velazquez, 2004) other argue that
either there is no convergence or, at best, the convergence is quite limited (Ferreiro et al
forthcoming; Starke et al, 2008)
The paper analyses whether there is a generalized change in the composition of public
expenditures towards those productive public expenditure in EU. With this aim, we
analyse the composition of public expenditure in the EU countries for the periods 199098 and after 1999. Using Tukey box-plots we analyse whether the composition of public
expenditures is converging in the EU countries and whether this convergence involve a
higher share of productive expenditures.
The paper structures as follows. The next section explores the theoretical basis of the
fiscal policy in the EU. Section 3 discusses the evolution of the composition of public
expenditure in EU countries, focusing the analysis on the evolution of those items that
are defined in the literature as productive public spending. Section 4 analyses the
convergence in the composition of public expenditures by means of a cluster analysis.
Final section summarizes and concludes.
2. The role of fiscal policy in the European Union: theoretical bases
The European Monetary Union has involved the implementation of a strategy of
macroeconomic policy that gives a special role to monetary policy, downgrading fiscal
policy to a secondary role, always subordinated to monetary policy. Fiscal policy is,
thereby, focused on the generation and maintenance of the right environment for the
effective working of monetary policy.
These theoretical foundations have led to a fiscal policy set by the rules and norms set
in the Maastricht Treaty and the Stability and Growth Pact. In this sense, the working of
national fiscal policies in the EU is based on the following principles:
3
i)
The implementation of sound and sustainable fiscal policies.
ii)
The reduction of the size of public deficits and the stocks of public debt.
iii)
The reduction of the size of public expenditure and taxation.
According to this approach, fiscal policy only affects economic activity in the short run.
However, fiscal policy does not generate effects on the long run, that is, it can not affect
the potential output. The long-term effects of fiscal policy would arise from the nonKeynesian effects of fiscal policy. One of the main conclusions of this approach is that
fiscal consolidation has an expansionary impact on the economic activity both in the
short-term and in the long-term (Afonso, 2001, 2006; Alesina and Perotti, 1995, 1997;
Alesina et al, 2002; Alesina, Perotti and Tavares, 1998; Briotti, 2004, 2005; European
Commission, Directorate-General for Economic and Financial Affairs, 2003, 2004,
2007; Giavazzi, Japelli and Pagano, 1999, 2000; Giavazzi and Pagano, 1990; Giudice,
G., Turrini, A. and in’t Veld, J., 2003, 2007; Hemming et al, 2002; Kumar, Leigh and
Plekhanov, 2007; McDermott and Wescott, 1996; van Aarle and Garretsen, 2003).
Thus, fiscal policy in the European Union plays a passive role, subordinated to the
monetary policy f the ECB. The objective of fiscal policy is the creation of a sound
economic environment via the reduction-removal of fiscal imbalances that gives rise to
low inflation rates and that helps to an effective working of the monetary policy of the
European Central Bank. Moreover, the counter-cyclical fiscal policy must be only based
on the working of built-in stabilizers, and, consequently, only cyclical fiscal deficit are
allowed, although with a size below 3% GDP1.
This view about the fiscal policy is based on the axiom that demand-side policies do not
affect the economic activity in the long-run, that is, that fiscal policy does not influence
the path of potential output. However, fiscal policy is currently gaining relevance in the
EU putting the issue of the quality of public finances at the core of fiscal policies
(Deroose and Kastrop, 2008; European Commission, Directorate-General for Economic
and Financial Affairs, 2008). The Lisbon Strategy and the current Broad Economic
Policy Guidelines (BEPG) accept the positive impact that the fiscal policy can have in
the long-run, both in terms of the level and the rate of growth of potential output. This
4
impact would come from the composition of public expenditure, not from the size of
public expenditures or revenues or from the fiscal balance.
These arguments are now being accepted by the European institutions2. For the
Commission and the ECOFIN Council, public budgets can contribute to foster
economic growth and employment through three channels: “supporting a stable
macroeconomic framework via sound public finances, making tax and benefit systems
more employment friendly and redirecting public expenditures towards physical and
human capital accumulation.” (Council of the European Union, 2001, p. 1). The Broad
Economic Policy Guidelines for the 2005-08 period states in the guideline three that
“To promote a growth- and employment-orientated and efficient allocation of resources
Member States should, without prejudice to guidelines on economic stability and
sustainability, re-direct the composition of public expenditure towards growthenhancing categories in line with Lisbon strategy (…) This can be achieved by
redirecting expenditure towards growth-enhancing categories such as research and
development (R&D), physical infrastructure, environmental friendly technologies,
human capital and knowledge” (European Commission, 2005, p. 41).
The reformed SGP states that the Commission’s Report that must evaluate the existence
of an excessive deficit “shall appropriately reflect developments in the medium-term
economic position (in particular potential growth, prevailing cyclical conditions, the
implementation of policies in the context of the Lisbon agenda and policies to foster
research and development and innovation) and developments in the medium-term
budgetary position (in particular, fiscal consolidation efforts in “good times”, debt
sustainability, public investment and the overall quality of public finances)”3.
The public-policy endogenous growth models (PPEGMs) are the theoretical basis of
this new fiscal policy strategy. These models are an extension of the endogenous growth
theory, according to which the economic growth is determined by the accumulation of
productive factors and by technical progress (Barro, 1990; Barro and Sala-i-Martin,
1995; Grossman and Helpman, 1991; Jones, 1995; Lucas, 1988; Rebelo, 1991; Romer,
1986, 1990). These models focus on the role that fiscal policy can play in enhancing
economic growth: shifting the revenue stance away from distortionary forms of taxation
and towards non-distortionary forms, and switching expenditures from unproductive to
5
productive forms are growth-enhancing (Angelopoulos et al, 2006; Aschauer, 1989;
Barro, 1990, 1991; Baier and Glomm, 2001; Barro and Sala-i-Martin, 1995; Devarajan
et al, 1996; Gemmel and Kneller, 2001; Gupta el al, 2005; King and Rebelo, 1990;
Kneller et al, 1999, 2001; Kocherlatoky and Yi, 1997; Romero de Avila and Strauch,
2003; Zagler and Dürnecker, 2003).
PPEG models define as ‘productive’ expenditures those that, by a complementing
private sector production and generating positive externalities to firms, have a positive
effect on the marginal productivity of capital and labour, and ‘unproductive’
expenditures would be those that give direct utility to households (Angelopoulos,
Economides and Kamman, 2006; Devarajan et al, 1996; European Commission,
Directorate-General for Economic and Financial Affairs, 2004).
Although the empirical evidence is mixed, for a number of studies ‘productive’
expenditures include the following outlays: public investment, R&D, education, active
labour market policies, health, defence, public order and general administrative costs,
transport and communication (Afonso et al 2005; Afonso and González Alegre, 2008;
Angelopoulos et al 2006; Aschauer, 1989; Atkinson and van den Noord, 2001; Barro,
1990, 1991; Bleaney et al, 2001; Bloom et al, 2001; Devarajan et al, 1996; Easterly and
Rebelo, 1993; Gemmel and Kneller, 2001; Kneller et al, 1999, 2001; Lamo and Strauch,
2002; Nourzad and Vrieze, 1995; Romero de Avila and Strauch, 2003; Sanchez-Robles,
1998; Thöne, 2003).
However, empirical studies show that the impact of public expenditures depends on the
kind of taxation (distortionary or a non-distortionary) that finances that expenditure
(Kneller et al, 1999, 2001; Kocherlatoky and Yi, 1997; Romero de Avila and Strauch,
2003). Another caveat is the existence of non-linear relations between economic growth
and the size of public spending (Cameron, 1978). Thus, although public spending may
have a positive impact on growth, the trend reverse expenditure exceeds a certain
threshold, being this limit different for each type of spending (European Commission,
Directorate-General for Economic and Financial Affairs, 2002; Gupta et al, 2005).
Finally, empirical studies face additional problems. First, national statistics on
government outlays are not fully comparable (Florio, 2001). Second, the theoretical
6
classification of public outlays into ‘productive’ or ‘unproductive’ is not available at
macroeconomic level. This forces to use data coming from national accounts, either on
the basis of the economic or the functional classification, and to assume that in each
case all the spending in a certain category is either productive or unproductive
(European Commission, Directorate-General for Economic and Financial Affairs,
2004). Finally, the current level of aggregation of public expenditures, both in the
functional-COFOG and the economic classifications, is too high for a right assessment
of the impact of specific outlays on economic growth (Deroose and Kastrop, 2008).
3. An analysis of the convergence in the size and composition of public expenditure
in EU countries
The objective of the next sections is to analyse whether EU countries have converged in
the composition of their public expenditures. With this aim, we have compared the
evolution of public expenditures between two sub-periods: 1990-98 (the sub-period
called “before” EMU) and 1999-07 (“after” EMU). In order to avoid business cycle
effects on public expenditures and expenditure composition we have calculated the
average size and the composition of public spending for each of the two sub-periods
analysed. Data on economic expenditures, both for the size of public expenditures (as a
percentage of GDP) and for the composition of public expenditures (COFOG and
economic classification) have been obtained form the Eurostat Government Finance
Statistics at the Eurostat website.
The first analysis of the convergence of public expenditures will be carried out using
statistical measures of dispersion and boxplots4. A convergence process in fiscal
variables involves that after EMU, the member states have a more similar percentage in
each item than that existing in the period before EMU. Therefore, we study whether the
dispersion of each item has diminished comparing to the period before EMU. We use in
our analysis two measures of dispersion: standard deviation and interquartile range. The
standard deviation is a good measure of dispersion when there are not outliers. The
interquartile range is a robust measure, since it is not affected by outliers. Standard
deviation explains the dispersion of the whole distribution. Interquartile range only
explains 50% of the cases. The information provided by both measures is completed
with boxplots.
7
3.1. Is the size of public expenditures converging in EU countries?
As figure 1 shows the size of public expenditures has fallen in 22 out of the 26 countries
analysed. This evolution is in accordance to the view, previously analysed, that a fall in
the size of public sector has a positive impact on economic growth. Nonetheless, this
fall dies not mean necessarily that public expenditures are converging in the European
Union.
Figure 1. Evolution of the size of public expenditures (%GDP)
se
65
60
fi
dk
sk
at
55
fr
it
nl
Before
hu
50
de
be
pl
cz
ro
45
uk
lu
lit
pt
gr
es
si
mt
40
ie
ee
cy
lv
35
30
30
35
40
45
50
55
60
65
After
Source: our calculations
To reach a conclusion about this behaviour, we have made a box-plot analysis. In this
sense, we can talk of convergence when we detect a lower standard deviation and a
lower interquartile range. Figure 2 shows clearly that public expenditures are
converging in the EU to a lower percentage of the respective domestic GDPs5.
8
Figure 2. Box plot of the size of public expenditures (%GDP)
Public expenditures
60,00
55,00
50,00
45,00
40,00
35,00
30,00
Before
After
3.2. Is the composition of public expenditures converging in EU countries?
The composition of public expenditures has been analysed in terms on the economic
and functional (GOFOG) classifications. In both cases, the source of data is the same:
the Eurostat Government Finance Statistics. Unfortunately, not all EU countries have
available data before 1999, and, consequently, these countries have been excluded from
the analysis. Furthermore, the period for which data are available in the two subperiods
(26 countries for the economic classification and 20 countries for the functional
classification) is not the same for all the countries6. Nonetheless, in as much we are
calculating the average for the two sub-periods this problem does not affect to the
outcome.
If EU countries had implemented the recommendations from the PPEGMs, the shares of
those public expenditures considered as productive expenditures should have increased,
and viceversa. Although the theoretical distinction between productive and
unproductive might be clear, however, as we mentioned in previous sections, the
definition of a specific public expenditure as productive or unproductive must be taken
with caveats. Nonetheless, based on theoretical and empirical studies, we have
considered as productive expenditures the following items:
-
in the functional classification: defence, public order and safety, economic
affairs
(includes
sectorial
R&D
and
transport
and
communication),
9
environmental protection, housing and community amenities, health and
education;
-
in the economic classification: capital transfers and gross capital formation
Table 1 and 2 shows the evolution of the different categories of public expenditures in
both kinds of classification. These show the change in the shares of each kind of
expenditure as a percentage of total public expenditure. As can be seen, the picture is
not clear and we can not pose the existence of a common pattern for all the EU
countries or even for a single country. Furthermore, the conclusion would be dependent
on the initial size/share of each kind of expenditure.
In any case, since our main interest focus on the hypothesis of a convergence in the
public expenditures composition, we have analysed in this section that pattern of
behaviour by means of statistical measures of dispersion and of boxplots (tables 3 and 4
and figures 3 and 4). The analysis of the boxplots and, mainly, of the standard
deviations (columns 5, 6 and 7 of tables 3 and 4) and the interquartile ranges (columns
8, 9 and 10 of tables 3 and 4) allows to reach a significant number of conclusions about
the convergence process that might be taken place in some items of public expenditure,
both in terms of the economic and the functional classifications.
10
Table 1. Evolution of the composition of public expenditures. Functional (COFOG) classification
General public services
Defence
Public order and safety
Economic affairs
INCREASE
Czech Republic, Malta, Finland
DECLINE
Belgica, Denmark, Germany, Estonia, Ireland, Greece, Spain, France, Italy,
Cyprus, Latvia, Netherland, Austria, Portugal, Sweden, United Kingdom
Estonia, Italy, Latvia
Belgium, Czech Republic, Germany, Ireland, Greece, Spain, France, Cyprus,
Luxembourg, Malta, Portugal, Finland, Sweden, United Kingdom
Belgium, Denmark, Germany, Ireland, Greece, Spain, France, Italy,
Luxembourg,Netherland, Austria
Estonia, Cyprus, Latvia, Malta, Portugal, Finland, Sweden, United Kingdom
Belgium, Estonia, Ireland, Greece, Spain, Netherlands, Austria, Sweden
Czech Republic, Denmark, Germany, France, Italy, Luxembourg, Malta,
Portugal, Finland, United Kingdom, Cyprus, Latvia
Czech Republic, Germany, Malta, Austria, Luxembourg
Enviroment protection
United Kingdom, Cyprus, Latvia, Sweden, Finland, Belgium, Denmark, Estonia, Ireland,
Greece, Spain, France, Italy, Netherland, Portugal
Housing and comunities
amenities
Belgium, Germany, Ireland, Greece, Spain, France, Italy, Latvia
United Kingdom, Sweden, Finland, Czech Republic, Denmark, Estonia, Cyprus,
Luxembourg, Malta, Netherland, Austria, Portugal
Finland, Sweden, United Kingdom, Belgium, Czech Republic, Denmark, Germany,
Ireland, Greece, Spain, Italy, Latvia, Malta, Netherland, Portugal, France
Estonia, Cyprus, Luxembourg, Austria
United Kingdom, Finland, Belgium, Czech Republic, Denmark, Estonia, Ireland, Greece,
Spain, France, Italy, Latvia, Portugal, Luxembourg
Sweden, Germany
Portugal, Austria, Finland, Sweden, United Kingdom, Belgium, Czech Republic,
Denmark, Germany, Ireland, Spain, Italy, Cyprus, Latvia, Luxembourg, Netherlands
Sweden, Belgium, Czech Republic, Denmark, Germany, Estonia, France, Italy, Cyprus,
Luxembourg, Austria, Portugal
Estonia, Greece, France, Malta
Health
Recreation, culture and
religion
Education
Social protection
Finland, United Kingdom, Ireland, Greece, Spain, Latvia, Malta, Netherlands
Source: our calculations
11
Table 2. Evolution of the composition of public expenditures. Economic classification
Intermediate consumption
INCREASE
Belgium, Czech Republic, Denmark, Germany, Ireland, Greece, Spain, Italy, Luxembourg,
Netherlands, Poland, Romania, Finland, Sweden, United Kingdom
DECLINE
Estonia, France, Cyprus, Latvia, Lithuania, Hungary, Malta, Austria, Slovenia, Slovakia
Belgica, Czech Republic, Denmark, Estonia, Ireland, Greece, Spain, France, Italy, Latvia,
Lithuania, Hungary, Netherlands, Poland, Portugal, Romania, Slovenia, Slovakia, Sweden
Germany, Cyprus , Luxembourg, Malta, Austria, United Kingdom
Belgium, Denmark, Estonia, Spain, Malta , Slovenia, Slovakia, Sweden
Czech Republic, Germany, Ireland, Greece, France, Italy, Cyprus, Latvia, Lithuania,
Luxembourg, Hungary, Netherlands, Slovenia, Slovakia, Sweden
Belgium, Lithuania, Malta, Slovakia,
Czech Republic, Denmark, Germany, Estonia, Ireland, Greece, Spain, France, Italy, Cyprus,
Latvia, Luxembourg, Hungary, Netherlands, Austria, Poland, Portugal, Romania, Slovenia,
Finland, Sweden, United Kingdom
Social benefits other than social
transfers in kind
Belgium, Czech Republic, Germany, Estonia, Greece, Italy, Cyprus, Hungary, Austria,
Poland, Portugal, Romania, Slovakia, United Kingdom, Lithuania, Luxembourg
Denmark, Ireland, Spain, France, Latvia, Malta, Netherlands, Finland, Sweden, United
Kingdom
Social transfers in kind
Belgium, Czech Republic, Denmark, Germany, Estonia, Ireland, Spain, France, Italy,
Latvia, Cyprus, Hungary, Luxembourg, Slovenia, Slovakia, Finland, Sweden, Romania,
Portugal, Poland, Hungary, Malta, Netherlands, Austria
Lithuania
Other current transfer
Belgium, Czech Republic, Denmark, Estonia, Ireland, Greece, Spain, France, Italy, Cyprus,
Latvia, Lithuania, Luxembourg, Hungary, Malta, Netherlands, Austria, Poland, Portugal,
Slovenia, Slovakia, Finland, Sweden, United Kingdom
Germany, Romania
Belgium, Estonia, Ireland, Italy, Cyprus, Latvia, Luxembourg, Malta, Austria
Czech Republic, Denmark, Germany, Greece, Spain, France, Lithuania, Hungary,
Netherlands, Poland, Postugal, Romania, Slovenia, Slovakia, Finland, Sweden, United
Kingdom
Belgium, Denmark, Ireland, Greece, Spain, Latvia, Lithuania, Hungary, Malta, Netherlands,
Poland, Romania, Slovenia, Sweden
Czech Republic, Germany, Estonia, France, Italy, Cyprus, Luxembourg, Austria, Portugal,
Slovakia, Finland, United Kingdom
Compensation of employees
Subsidies
Property income
Capital transfers
Gross capital formation and
adquisitions less disposals of nonfinancial assets
Source: our calculations
12
Table 3. Measures of dispersion of the composition of public expenditures. Economic classification
Intermediate consumption
Compensation of employees
Subsidies
Property income
Social benefits other than sbk
Social benefits in kind
Other current transfers
Capital transfers
Gross capital formation
Mean
after (1) before (2)
13.75
12.83
25.48
3.04
5.88
30.93
6.65
4.53
2.81
6.61
25.24
3.52
9.29
30.50
5.23
3.34
3.30
6.56
Median
after (3) before (4)
12.64
11.89
25.65
2.83
5.83
30.42
5.28
4.40
2.77
6.72
25.29
3.14
8.13
30.87
3.44
3.44
2.83
5.91
Standard deviation (7=5/6)*100
after (5) before (6)
4.91
5.14
95.66
5.08
1.39
3.26
4.08
5.48
1.50
1.68
2.87
5.13
1.41
6.17
3.78
4.76
1.31
2.66
2.47
99.05
98.47
52.87
108.09
115.07
114.10
63.03
115.94
Interquartile range
after (8) before (9)
7.80
5.48
7.32
1.50
3.73
6.94
9.95
2.28
2.32
5.46
5.92
2.48
7.38
5.71
8.35
1.75
2.00
3.82
(10=8/9)
142.23
123.62
60.20
50.49
121.59
119.21
130.28
115.85
142.76
Source: our calculations
Table 4. Measures of dispersion of the composition of public expenditures. Functional classification
General public services
Defence
Public order and safety
Economic affairs
Environmental protection
Housing
Health
Recreation
Education
Social protection
Mean
after (1) before (2)
14.77
17.60
3.27
3.59
3.86
3.70
10.58
11.30
1.55
1.44
2.10
2.44
12.78
10.96
2.61
2.30
12.58
11.89
35.89
34.80
Median
Standard deviation
after (3) before (4) after (5) before (6)
14.22
16.82
3.67
6.22
3.11
3.25
1.48
1.73
3.65
3.58
1.40
1.70
10.64
10.95
3.05
3.99
1.53
1.30
0.64
0.76
1.98
2.13
1.13
1.38
13.11
11.73
2.71
2.54
2.44
2.25
1.21
1.07
12.05
11.45
2.89
2.79
36.95
34.57
6.18
6.05
(7=5/6)*100
58.95
85.61
82.31
76.50
84.53
82.22
106.71
112.96
103.50
102.20
Interquartile range
after (8) before (9)
4.90
7.75
1.44
1.69
2.43
2.63
3.63
3.81
0.97
0.91
1.30
1.27
2.09
2.35
1.24
1.04
3.53
2.76
10.86
10.03
(10=8/9)
63.23
84.82
92.36
95.31
106.55
102.64
88.99
119.15
127.72
108.21
Source: our calculations
13
Figure 3. Box-plots of the dispersion of public expenditures. Economic classification (in
Estonia
25
Latvia
20
15
10
5
Before
6
5
4
3
2
1
Cyprus
35
30
25
20
15
Before
0
25
After
Greece
20
15
10
5
0
Before
After
Before
After
Before
After
20
Social tranfers in kind
Social benefits other than
social transfers in kind
40
After
Austria
7
subsidies payable
Compensation of employees
30
property income, payable
Intermediate Consumption
percentage of total public expenditure)
40
35
30
25
15
10
5
0
Before
After
14
14
7
Capital tranfers,payable
other current transfers
8
Luxembourg
6
5
4
3
2
Czech Republ
12
10
8
Czech Republ
6
4
2
0
1
Before
Before
After
After
Gross capital formation
12
10
8
6
4
2
Before
After
Figure 4. Box-plots of the dispersion of public expenditures. COFOG classification (in
percentage of total public expenditure)
33
8
Greece
30
Greece
United Kingdom Greece
United Kingdom
6
27
Cyprus
24
21
18
Defence
General Public services
36
4
2
15
12
9
0
Before
After
Before
After
15
6
4
2
0
Czech Republic
15
10
After
Luxembourg
Before
Housing and community
amenities
Environment protection
20
5
Before
3
2
1
0
Cyprus
5
Malta
Cyprus
4
3
2
1
After
15
12
9
Latvia
Greece
Cyprus
7
Recreation, culture and
religion
18
6
Before
Ireland
21
Health
6
After
0
Before
3
20
6
After
Estonia
Estonia
5
4
3
2
1
0
Before
After
Before
After
50
Estonia
Social protection
18
16
Education
Czech Republ
25
Economic affairs
Public order and safety
8
14
12
10
45
40
35
30
25
8
Greece
20
6
Before
After
Before
After
16
If we identify convergence with a reduction in standard deviation, in interquartile range
and in the upper and lower adjacent values, we can detect in terms of the economic
classification a convergence process to lower shares in the items of subsidies and
property income (to lower shares) and a strong divergence in social transfers in kind and
other current transfers (to higher shares), in gross capital formation (to similar shares)
and in social benefits other than social transfers in kind (although in this case the shares
are stable). In the case of intermediate consumption and compensation of employees,
the standard deviation falls because the outliers detected disappear: once considered this
behaviour, we detect a divergence process to higher shares. Finally in the case of capital
transfers, the ruling out of the far outlier hides a diverging process to higher shares of
this item.
In terms of the COFOG classification a strong convergence to lower shares is detected
in general public services, defence and economic affairs, and to higher shares in public
order and safety. On the contrary, a diverging process to higher shares is detected in
recreation, culture and religion, education and social protection. The analysis of the
other items is less clear due to the existence of outliers. Nonetheless, the box-plots show
a divergence to higher shares in the case of health, and a convergence to lower shares in
housing and community amenities and a convergence to higher shares in environment
protection.
To conclude, the empirical evidence of a convergence process of the productive items of
public expenditure is quite mixed. Having followed the recommendations of the
PPEGMs, EU countries should have converged in the composition of public
expenditures with an increase in the shares of this kind of expenditures. However, only
three categories of productive public expenditures have increased their shares: public
order and safety, health and education, and in the two latter a diverging process is
detected, sign of a different pace of evolution. Actually, as the box-plots show, only in
the case of education a generalized rise is detected, whilst in health the convergence is
explained by the increase in the lower adjacent value and the increase in the main is
mainly explained by the appearance of a far outlier with the highest share (Ireland).
17
4. A cluster analysis of the composition of public expenditures in EU Member
States
The previous analysis is now complemented with a cluster analysis of the components
of public spending using the economic and the COFOG classification. The existence of
a convergence process should have lead to a lower number of clusters. To elaborate the
clusters and to study their evolution before and after 1999, we have used a cluster
analysis, in particular the Ward’s criterion (Ward, 1963). To identify cluster of cases we
have calculated factor coordinates. Principal component analysis (PCA) gives the more
explicative factor of the whole set of countries. An ascending hierarchical classification
has been applied to coordinates estimated by PCA using Ward’s criterion7. Test values8
show the main features of each cluster (using 5% as critical probability frontier9). In the
cluster analysis of the economic classification, the data matrix is formed by 20 countries
and 10 active variables: the shares as percentage of total public expenditure of the 9
items in this classification plus the size of public expenditure (as percentage of GDP). In
the cluster analysis of the COFOG classification, the data matrix is formed by 20
countries and 11 active variables: the shares as percentage of total public expenditure of
the 10 items in this classification plus the size of public expenditure (as percentage of
GDP).
18
Table 5. Clusters of EU countries according to the composition of public expenditures
Economic classification
Before EMU
Countries
Significant shares above
average
Significant shares below
average
After EMU
Countries
Significant shares above
average
Cluster 1
Cluster 2
Austria, Denmark, Germany, France, Belgium, Greece, Italy, Ireland,
Finland, Netherlands, Luxembourg, Spain
Sweden
Size public expenditure
Property income
Social benefits other than stk
Subsidies
Social tranfers in kind
Subsidies
Cluster 3
Czech Republic
Cluster 4
Cyprus, Malta, Portugal
Cluster 5
Estonia, Latvia, United Kingdom
Capital transfers
Compensation of employees
Intermediate consumption
Property income Subsidies
Size of public expenditure
Cluster 1
Austria, Czech Republic, Germany,
Luxembourg
Size public expenditure
Social benefits other than stk
Subsidies
Social tranfers in kind
Cluster 2
Belgium, France, Greece, Italy,
Malta, Netherlands, Portugal Spain
Property income
Cluster 3
Cyprus, Denmark, Finland, Sweden,
United Kingdom
Intermediate consumption
Compensation of employees
Gross capital formation
Social benefits other than stk
Cluster 4
Estonia, Ireland, Latvia
Other current transfers
Intermediate consumption
Capital transfers
Social transfers in kind
Significant shares below
average
Gross capital formation Intermediate
consumption
Social benefits other than stk
Property income
Size of public expenditure
COFOG classification
Before EMU
Countries
Significant shares
above average
Cluster 1
Denmark, Finland, France,
Netherlands, Sweden,
United Kingdom
Size public expenditures
Social protection
Significant shares
below average
After EMU
Countries
Environment protection
Economic affairs
Cluster 1
Estonia, Latvia, Portugal,
United Kingdom
Significant shares
above average
Public order and safety
Education
Recreation
Size public expenditure
Significant shares
below average
Cluster 2
Austria, Germany,
Luxembourg
Cluster 3
Belgium, Ireland, Italy,
Portugal, Sweden
Greece
Environment protection
Social protection
Cluster 4
Cluster 5
Czech Republic, Malta
General public services
General public services
Defence
Economic affairs
Environment protection
Social protection
Health
Cluster 2
Czech Republic, Ireland,
Luxembourg, Malta, Spain
Cyprus
Cluster 3
Greece
Economic affairs
Environment protection
Health
Defence
Size public expenditures
General public services
Housing and community
amenities
Health
Social protection
Defence
Cluster 6
Cyprus
Housing s
Social protection
Cluster 4
Education
Cluster 7
Estonia, Latvia
Education
Public order and safety
Recreation
Size public expenditures
Cluster 5
Austria, Belgium,
Denmark, Finland, France,
Germany, Italy,
Netherlands, Sweden
Size public expenditures
Social protection
Environment protection
Public order and safety
Economic affairs
Source: Our calculations
19
Table 6. Relevant expenditures in the formation of clusters. Economic classification
BEFORE EMU
Clusters
Variables
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
Cluster
average
Overall
average
Cluster
standard
deviation
Overall
standard
deviation
Test Value
Probability
Size public expenditures
52.90
46.89
4.77
7.07
3.03
0.001
Social benefits other than
stk
33.16
30.50
1.88
3.68
2.57
0.005
Subsidies
Social tranfers in kind
Property income
Subsidies
Capital transfers
Compensation of
employees
4.38
7.74
17.33
2.29
13.51
33.78
3.52
5.23
9.29
3.52
3.30
25.24
0.87
4.55
4.25
0.69
3.04
1.37
4.64
6.02
1.37
2.66
5.00
2.23
1.93
3.36
-2.26
3.84
3.13
0.013
0.027
0.000
0.012
0.000
0.001
Social benefits other than
stk
25.73
30.50
2.33
3.68
-2.37
0.009
Intermediate consumption
22.92
12.83
2.15
5.01
3.69
0.000
Property income
Subsidies
Size public expenditures
3.73
2.04
39.06
9.29
3.52
46.89
2.99
0.46
3.67
6.02
1.37
7.07
-1.69
-1.97
-2.03
0.045
0.024
0.021
Cluster
average
Overall
average
Cluster
standard
deviation
Overall
standard
deviation
Test Value
Probability
4.70
4.66
12.62
3.04
2.81
6.65
1.38
1.65
2.10
1.35
1.63
5.34
2.68
2.47
2.44
0.004
0.007
0.007
34.93
30.93
4.45
3.98
2.19
0.014
18.13
25.48
1.29
4.95
-3.23
0.001
8.49
3.76
5.88
4.53
2.52
0.73
3.18
1.46
2.92
-1.86
0.002
0.031
10.96
13.75
2.27
4.79
-2.07
0.019
17.79
13.75
4.16
4.79
2.12
0.017
29.47
25.48
3.57
4.95
2.02
0.021
2.34
1.23
10.05
6.65
2.81
6.61
2.03
0.65
1.66
5.34
1.63
2.80
-2.03
-2.44
2.25
0.021
0.007
0.012
19.11
13.75
2.56
4.79
2.05
0.020
26.64
30.93
0.60
3.98
-1.97
0.024
2.18
5.88
1.46
3.18
-2.13
0.017
35.21
44.65
1.18
5.74
-3.01
0.001
AFTER EMU
Clusters
Variables
Cluster 1
Subsidies
Capital transfers
Social tranfers in kind
Social benefits other than
stk
Cluster 2
Compensation of
employees
Property income
Other current transfers
Intermediate consumption
Cluster 3
Intermediate consumption
Compensation of
employees
Social tranfers in kind
Capital transfers
Cluster 4
Gross capital formation
Intermediate consumption
Social benefits other than
stk
Property income
Size public expenditures
Source: our calculations
20
Table 7. Relevant expenditures in the formation of clusters. COFOG classification
BEFORE EMU
Clusters
Variables
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
Cluster 6
Cluster 7
Cluster
average
Overall
average
Cluster
standard
deviation
Overall
standard
deviation
Test Value
Probability
Size public expenditures
53.48
46.89
5.03
7.07
2.66
0.004
Social protection
Environment protection
Economic affairs
Environment protection
Social protection
General Public services
Health
Social protection
General public services
Defence
Health
Economic affairs
Environment protection
Housing
Social protection
Education
Public order and safety
Recreation
Size public expenditures
40.00
0.96
8.53
2.23
40.80
21.37
12.24
31.33
35.32
7.54
4.96
20.69
2.67
5.55
20.89
17.57
6.87
4.00
36.93
34.80
1.44
11.30
1.44
34.80
17.60
10.96
34.80
17.60
3.59
10.96
11.30
1.44
2.44
34.80
11.89
3.70
2.30
46.89
1.54
0.47
1.95
0.67
0.97
3.13
1.33
3.11
1.20
0.23
1.47
2.55
5.90
0.74
3.88
0.74
5.90
6.07
2.47
5.90
6.07
1.73
2.54
3.88
0.74
1.38
5.90
2.72
1.66
1.04
7.07
2.52
-1.88
-2.04
1.95
1.86
1.56
1.30
-1.48
2.85
2.28
-2.36
3.51
2.42
2.26
-2.30
3.03
2.78
2.36
-2.05
0.006
0.030
0.021
0.025
0.031
0.059
0.097
0.069
0.002
0.011
0.009
0.000
0.008
0.012
0.011
0.001
0.003
0.009
0.020
Cluster
average
Overall
average
Cluster
standard
deviation
Overall
standard
deviation
Test Value
Probability
3.54
0.19
AFTER EMU
Clusters
Variables
Cluster 1
Public order and safety
Education
Recreation
Size public expenditures
5.74
16.21
3.58
39.79
3.86
12.58
2.61
44.65
1.01
1.92
1.54
4.05
1.36
2.81
1.18
5.74
3.00
2.81
1.80
-1.85
0.001
0.002
0.036
0.032
Cluster 2
Economic affairs
Environment protection
Health
Defence
Size public expenditures
14.20
2.27
14.65
2.14
40.13
10.58
1.55
12.78
3.27
44.65
2.58
0.36
2.92
1.00
3.82
2.97
0.62
2.64
1.45
5.74
3.06
2.90
1.78
-1.97
-1.98
0.001
0.002
0.038
0.024
0.024
Cluster 3
General public services
Housing
Health
Social protection
Defence
Education
Size public expenditures
23.16
5.51
7.35
22.25
6.84
6.41
49.70
14.77
2.10
12.78
35.89
3.27
15.28
44.65
3.02
3.67
1.13
2.71
6.18
1.48
2.89
5.74
2.29
3.00
-2.00
-2.21
2.41
-2.13
3.47
0.011
0.001
0.023
0.014
0.008
0.016
0.000
Social protection
Environment protection
Public order and safety
Economic affairs
40.41
1.18
2.97
8.63
35.89
1.55
3.86
10.58
3.25
0.45
0.65
1.47
6.03
0.62
1.36
2.97
2.96
-2.37
-2.58
-2.58
0.002
0.009
0.005
0.005
Cluster 4
Cluster 5
Source: our calculations
21
Table 5, 6 and 7 show the results of the cluster analysis. The different variables that
appear in each cluster are ranked in terms of their relevance (test values) in the
formation of the clusters. The variables that contribute to the formation of the clusters
are those with a probability lower than 5%.
In the analysis of the economic classification of public expenditures before EMU, the
first five factors10 explain 89.1% of total inertia, with all countries being well explained
by these 5 factors. Ward’s criterion has been applied to these four factors. Using three
criteria (the ratio inertia inter/inertia total11, the structure of dendograms and the
significance of the classes) we have obtained five clusters. In the analysis of the data
after EMU, we have kept 5 factors, explaining 86.2% of total inertia, with all countries
being well represented by these factors. The Ward’s criterion has been used getting 4
clusters
The cluster analysis reinforces the conclusion reached in previous section. Thus, the
convergence in the size of public expenditures makes this variable less relevant in the
formation of the clusters, and after EMU only matters for cluster 4 (Estonia, Ireland and
Latvia) instead of the two countries (1 and 5) and 12 countries in which it was relevant
in the previous period. Furthermore, items where we detected a divergence process are
now more relevant affecting a higher numbers of clusters and/or clusters: intermediate
consumption (cluster 5 vs clusters 2-3-4), social transfers in kind (cluster 1 vs clusters
1-3), other current transfers (now in cluster 2), compensation of employees (cluster 4 vs
cluster 3) and capital transfers (clusters 3). Finally, we must stress that all items matter
in the formation of clusters in both sub-periods, what involves the maintenance of
significant differences in the respective shares in public budget. In sum, the cluster
analysis shows that there is no significant process in the composition of public
expenditures in the EU and, also, that there is no convergence process in the productive
expenditures.
In the analysis of the COFOG classification, before EMU, the first five factors explain
84.7%% of total inertia, with all countries being well explained by these 5 factors.
22
Ward’s criterion has been applied to these factors. Using the same three criteria three
criteria we have obtained seven clusters. In the analysis of the data after EMU, we have
kept 5 factors, explaining 86.4% of total inertia, with all countries being well
represented by these factors. The Ward’s criterion has been used getting 5 clusters
Again, the cluster analysis reinforces previous conclusions. The convergence in some
outlays leads to a lower number of clusters; moreover, the higher convergence in the
composition of public expenditures (in comparison with that detected in the economic
classification) makes the size of public expenditures a more relevant variables,
influencing in the sub-period after EMU in three clusters (1, 2 and 5) instead of the 2
clusters before EMU. In any case, like in the economic classification, all items matter in
the formation of clusters in both sub-periods, what involves the maintenance of
significant differences in the respective shares of public expenditure. In sum, the cluster
analysis shows that there is no significant process in the composition of public
expenditures in the European Union.
5. Summary and conclusions
The current strategy of fiscal policy in the EU is changing in recent years. Fiscal
policies pay a higher attention to the quality of public finance and to the composition of
public expenditures. Having followed these principles, the EU Member States should
have increased the share of those components than can be defined as productive
expenditures.
The analysis made in the paper shows that, in practice, national fiscal policies have only
converged in the size of public expenditures with a generalized fall in the size of public
budget. However, differences in the composition of public spending remain both if we
analyse the economic or the functional classification, although a higher convergence is
detected in the case of the COFOG classification. In any case, the analysis do not show
an generalized increase or and upwards convergence in the productive components of
public expenditures.
Biographical note
23
Jesus Ferreiro is Associate Professor of Economics at the Department of Applied
Economics V of the University of the Basque Country, Spain. M. Teresa Garcia-delValle is Associate Professor of Statistics at the Department of Applied Economics V of
the University of the Basque Country, Spain. Carmen Gomez is Lecturer of Economics
at the Department of Applied Economics V of the University of the Basque Country,
Spain.
Address for correspondence
Jesus Ferreiro
Department Applied Economics V
Faculty of Economics and Business
University of the Basque Country
Avenida Lehendakari Agirre, 83
48015 Bilbao
Spain
E-mail: [email protected]
Notes
1. See European Commission, Directorate-General for Economic and Financial Affairs
of European Commission and European Central Bank (2006, 2004a, 2004b, 2001).
2. The ECB also stresses the need to combine the measures of consolidation of fiscal
imbalances with the improvement in the quality of public finances, changing the
composition of public spending towards productive expenditures (European Central
Bank, 2008a, 2008b).
3. Article 1.3 Council Regulation (EC) No 1467/97 of 7 July 1997, as amended by
Council Regulation (EC) No. 1056/05 of 27 June 2005 (ithalics added).
4. A boxplot (Tukey, 1977) summarizes the distribution of a set of data by displaying
the centering and spread of data using five elements: the smallest observation, the first
quartile, the third quartile, the media and the largest observation. First and third
quartiles are termed the hinges, and the difference between them represents the
interquartile range (IQR). Median is depicted using a line through the center of the box.
24
The inner fence is defined as the first quartile minus 1.5*IQR and the third quartile plus
1.5*IQR. Whiskers and staple show the values outside the first (lower adjacent value)
and third quartiles (upper adjacent value) but within the inner fences. The staple is a line
drawn at the last data point within (or equal) each of the inner fences. Whiskers are
horizontal lines drawn from each hinge to the corresponding staple. Any data lying
more than 1.5*IQR lower than the first quartile or 1.5*IQR higher than the third quartile
is considered as outlier. To characterize outliers, the outer fence is defined as the first
quartile minus 3.0*IQR and the third quartile plus 3.0*IQR. Data between the inner and
outer fenced are named near outliers (circles), and data outside the outer fence (stars)
are far outliers.
5. The mean of public expenditures (as a percentage of GDP) falls from 46.89% to
44.65%; the median falls from 45.24% to 44.90%, the standard deviation falls from 7.25
to 5.89 and the interquartile range falls from 10.79 to 8.72.
6. In the functional classification, available dates are: 1990-2007 (Denmark and
Luxembourg), 1990-2006 (Belgium, Ireland, Greece, Italy, Portugal, Finland, United
Kingdom), 1991-2007 (Germany), 1995-2006 (Czech Republic, Estonia, Spain, Malta,
Netherlands, Austria, Sweden), 1995-2006 (France) and 1998-2006 (Cyprus, Latvia). In
the economic classification, available dates are: 1990-2007 (Austria, Belgium,
Denmark, Germany, Greece, France, Finland, Italy, Luxembourg, Netherlands,
Portugal, United Kingdom), 1993-2007 (Latvia, Slovakia, Sweden) 1995-2007 (Czech
Republic, Estonia, Ireland, Lithuania, Malta, Poland, Spain, Slovenia), 1996-2007
(Hungary), 1998-2007 (Cyprus, Romania)
7. This criterion maximizes the variance among clusters and minimizes the variance
within clusters.
8. Test value (Lebart et al, 1995)) is a descriptive index used in the correspondence
analysis built following the methodology of hypothesis tests. A variable is not a relevant
to form a group if the nk values of this variable seem to have been randomly extracted
among the n observed values. The more doubtful the hypothesis of a random extraction,
the best this variable will characterize the group of nk individuals. For the categories of
nominal variables, the order allows to get those categories whose proportion within the
group is sufficiently different from the overall proportion, because it is higher (a
positive test value) or lower (a negative test value). Critical probabilities are used to
classify the variables by order of relevance. The most relevant or critical variable is that
with the lowest probability. In the study we use 5% as the frontier of critical probability.
25
9. The exception is cluster 3 before EMU in the COFOG classification, where 10% is
used as critical frontier.
10. The factors are linear combination of the initial variables.
11. The value of this ratio is 0.6456 before EMU and 0.525 after EMU in the economic
classification. In the COFOG classification these values are 0.7718 and 0.6856,
respectively
References
Afonso, A. (2001) ‘Non-Keynesian effects of fiscal policy in the EU-15’, Working
Paper, Department of Economics, Institute for Economics and Business Administration
(ISEG), Technical University of Lisbon, WP 2001/07.
Afonso, A. (2006) ‘Expansionary fiscal consolidations in Europe’, Working Paper
Series European Central Bank 675.
Afonso, A., Ebert, W., Schuknecht, L. and Thöne, M. (2005) ‘Quality of public finances
and growth’, Working Paper Series European Central Bank 438.
Afonso, A. and González Alegre, J. (2008) ‘Economic growth and budgetary
components. A panel assessment for the EU’, Working Paper Series European Central
Bank 848.
Alesina, A. and Perotti, R. (1995) ‘Fiscal expansions and adjustments in OECD
countries’, Economic Policy 21: 205-240.
Alesina, A. and Perotti, R. (1997) ‘Fiscal adjustments in OECD countries: composition
and macroeconomic effects’, IMF Staff Papers 44: 297-329.
Alesina, A., Ardagna, S., Perotti, R. y Schiantarelli, F. (2002) ‘Fiscal policy, profits and
investment’, American Economic Review 92: 571-589.
Alesina, A., Perotti, R. and Tavares, J. (1998) ‘The political economy of fiscal
adjustments’, Brooking Papers on Economic Activity 1: 197-248.
26
Angelopoulos, K., Economides, G. and Kamman, P. (2006) ‘Tax-spending policies and
economic growth: theoretical predictions and evidence from the OECD’, European
Journal of Political Economy doi: 10.1016/j.ejpoleco.2006.10.001.
Aschauer, D. (1989) ‘Is government spending productive?’, Journal of Monetary
Economics 23: 177-200.
Atkinson, P. and van den Noord, P. (2001) ‘Managing public expenditure: Some
emerging policy issues and a framework for analysis’, OECD Economics Department
Working Paper 285.
Baier, S. and Glomm, G. (2001) ‘Long-run growth and welfare effects of public policies
with distortionary taxation’, Journal of Economic Dynamics and Control 25: 20072042.
Barro, R.J. (1990) ‘Government spending in a simple model of endogenous growth’,
Journal of Political Economy 98(1): s103-117.
Barro, R.J. (1991) ‘Economic growth in a cross-section of countries’, Quarterly Journal
of Economics 106: 407-444.
Barro, R.J: and Sala-i-Martin, X. (1995) Economic Growth, New York: McGraw Hill.
Bleaney, M., Gemmell, N. and Kneller, R. (2001) ‘Testing the endogenous growth
model: public expenditure, taxation and growth over the long run’, Canadian Journal of
Economics 34(1): 36-57.
Bloom, D., Canning, D. and Sevilla, J. (2001) ‘The effect of health on economic
growth: theory and evidence’, NBER Working Paper Series 8587.
Briotti, M.G. (2004) ‘Fiscal adjustment between 1991 and 2002: stylised facts and
policy implications’, Occasional Paper Series, European Central Bank, 9.
27
Briotti, M.G. (2005) ‘Economic reactions to public finance consolidation: a survey of
the literature’, Occasional Paper Series, European Central Bank, 38.
Council of the European Union (2001) The Contribution of Public Finances to Growth
and Employment: Improving Quality and Sustainability, Report from the Commission
and the (ECOFIN) Council to the European Council (Stockholm, 23/24 2001), 6997/01,
Brussels, 12 March 2001.
Deroose, S. and Kastrop, C. (ed.) (2008) The Quality of Public Finances, European
Economy, Occasional Papers 37, Brussels: European Commission, Directorate-General
for Economic and Financial Affairs.
Devarajan, S., Swaroop, V. and Zou, H-F. (1996) ‘The composition of public
expenditure and economic growth’, Journal of Monetary Economics 37: 313-344.
Easterly, W. and Rebelo, S. (1993) ‘Fiscal policy and economic growth: An empirical
investigation’, Journal of Monetary Economics 32: 417-458.
European Central Bank (2001) ‘Fiscal policies and economic growth’, ECB Monthly
Bulletin January: 39-53.
European Central Bank (2004a) ‘Fiscal policy influences on macroeconomic stability
and prices’, ECB Monthly Bulletin April: 45-57.
European Central Bank (2004b) ‘EMU and the conduct of fiscal policies’, ECB Monthly
Bulletin January: 41-50.
European Central Bank (2006) ‘The importance of public expenditure reform for
economic growth and stability’, ECB Monthly Bulletin April: 61-73.
European Central Bank (2008a) Annual Report 2007, Frankfurt: European Central
Bank.
28
European Central Bank (2008b) ‘Economic and monetary developments’, ECB Monthly
Bulletin April: 9-56.
European Commission (2005) ‘The Broad Economic Policy Guidelines (for the 2005-08
Period)’, European Economy 4/2005.
European Commission, Directorate-General for Economic and Financial Affairs
(several years) ‘Public Finances in EMU’, European Economy.
Ferreiro, J., Garcia-del-Valle, M.T. and Gomez, C. (forthcoming) ‘Is the composition of
public expenditures converging in EMU countries?, Journal of Post Keynesian
Economics.
Florio, M. (2001) ‘On cross-country comparability of government statistics: public
expenditure trends in OECD national accounts’, International Review of Applied
Economics 15(2): 181-198.
Gemmel, N. and Kneller, R. (2001) ‘The impact of fiscal policy on long-run growth’,
European Economy. Reports and Studies 1: 97-129.
Giavazzi, F. and Pagano, M. (1990) ‘Can severe fiscal contractions be expansionary’
Tales of two small European countries’, NBER Macroeconomics Annual 5: 75-111.
Giavazzi, F., Japelli, T. and Pagano, M. (1999) ‘Searching for non-Keynesian effects of
fiscal policy’, Working Paper Centro Studi in Economia e Finanza 16.
Giavazzi, F., Japelli, T. and Pagano, M. (2000) ‘Searching for the non-linear effects of
fiscal policy: Evidence for industrial and developing countries’, NBER Working Paper
Series 7460.
Giudice, G., Turrini, A. and in’t Veld, J. (2003) ‘Can fiscal consolidation be
expansionary in the EU? Ex-post evidence and ex-ante analysis’, Economic Papers 195,
European Commission, Directorate-General for Economic and Financial Affairs.
29
Giudice, G., Turrini, A. and in’t Veld, J. (2007) ‘Non-Keynesian fiscal adjustments? A
close look at expansionary fiscal consolidations in the EU’, Open Economics Review
18(5): 613-630.
Grossman, G.M. and Helpman, E. (1991) Innovation and Growth in the Global
Economy, Cambridge, MA.: The MIT Press.
Gupta, S., Clements, B., Baldacci, E. and Mulas-Granados, C. (2005) ‘Fiscal policy,
expenditure composition and growth in low-income countries’, Journal of International
Money and Finance 24: 441-463.
Hemming, R., Kell, M. and Mahfouz, S. (2002) ‘The effectiveness of fiscal policy in
stimulating economic activity. A review of the literature’, IMF Working Paper
WP/02/208.
Jones, C.I. (1995) ‘R&D models of economic growth’, Journal of Political Economy
103: 759-784.
King, R.G. and Rebelo, S. (1990) ‘Public policy and economic growth: developing
neoclassical implications’, NBER Working Paper 3338.
Kneller, R., Bleaney, M.F. and Gemmell, N. (1999) ‘Fiscal policy and growth: evidence
from OECD countries’, Journal of Public Economics 74: 171-190.
Kneller, R., Bleaney, M.F. and Gemmell, N. (2001) ‘Testing the endogenous growth
model: public expenditure, taxation and growth over the long-run’, Canadian Journal of
Economics 34: 36-57.
Kocherlatoky, N. and Yi, K.M. (1997) ‘Is there endogenous long-term growth?
Evidence from the U.S. and the U.K.’, Journal of Money, Credit and Banking May:
235-262.
Kumar, M.S., Leigh, D. and Plekhanov, A. (2007) ‘Fiscal adjustments: determinants
and macroeconomic consequences’, IMF Working Paper WP/07/178.
30
Lamo, A. and Strauch, R. (2002) ‘The contribution of public finances to the European
growth strategy’, in Banca d’Italia, The Impact of Fiscal Policy, Rome: Banca d’Italia,
479-519.
Lucas, R.E. (1988) ‘On the mechanics of economic development’, Journal of Monetary
Economics 22: 3-42.
McDermott, C.J. and Wescott, R.F. (1996) ‘An empirical analysis of fiscal
adjustments’, IMF Staff Papers 43: 725-753.
Nourzad, F. and Vrieze, M.D. (1995) ‘Public capital formation and productivity growth:
some international evidence’, The Journal of Productivity Analysis 6: 283-295.
Rebelo, S. (1991) ‘Long-run policy analysis and long-run growth’, Journal of Political
Economy 99: 500-521.
Romer, P.M. (1986) ‘Increasing returns and long run growth’, Journal of Political
Economy 94: 1002-1037.
Romer, P.M. (1990) ‘Endogenous technical change’, Journal of Political Economy 98:
S71-S102.
Romero de Avila, D. and Strauch, R. (2003) ‘Public finances and long-term growth in
Europe: Evidence from a panel data analysis’, ECB Working Paper 246.
Sanchez- Robles, B. (1998) ‘Infrastructure investment and growth: some empirical
evidence’, Contemporary Economic Policy 16: 98-108.
Sanz, I and Velazquez, F.J. (2004) ‘The evolution and convergence of the government
expenditure composition in the OECD countries’, Public Choice 119: 61-72.
31
Starke, P., Obinger, H. and Castles, F.G. (2008) ‘Convergence towards where: in what
ways, if any, are welfare states becoming more similar’, Journal of European Public
Policy 15(7): 975-1000.
Thöne, M. (2003) ‘Public expenditure for growth and sustainable development (PEGS):
Conceptual, empirical and quantitative issues’, summary and main conclusions of a
study commissioned by the German Federal Ministry of Finance, Cologne Centre for
Public Finance.
Tukey, J.W. (1977) Exploratory Data Analysis, Reading: Addison-Wesley.
Van Aarle, B. and Garretsen, H. (2003) ‘Keynesian, non-Keynesian or no effects of
fiscal policy changes? The EMU case’, Journal of Macroeconomics 25(2): 213-240.
Ward, J.H. (1963) ‘Hierarchical groupings to optimize an objective function’, Journal
of the American Statistical Association 58: 236-244.
Zagler, M. and Dürnecker, G. (2003) ‘Fiscal policy and economic growth’, Journal of
Economic Surveys 17(3): 397-418.
32